Genetic variants of UDP‐glucuronosyltransferase 1A genes are associated with disease presentation and outcome in primary sclerosing cholangitis

Primary sclerosing cholangitis (PSC) is a progressive cholestatic liver disease without a curative medical therapy. The human UDP‐glucuronosyltransferases 1A play a major role in the detoxification and elimination of bilirubin, bile acids and xenobiotics. Whether genetic UGT1A variants determine course and outcome of PSC has not yet been described.


| INTRODUC TI ON
Primary sclerosing cholangitis (PSC) is a chronic progressive cholestatic liver disease with a strong clinical association with inflammatory bowel diseases (IBD). Patients with PSC suffer from recurrent episodes of cholangitis resulting in biliary cirrhosis, and are at a high risk to develop hepatobiliary or colorectal malignancies. 1 At present, there is no curative medical therapy available and therefore liver transplantation (LT) remains the only definitive treatment. 2 The pathogenesis of PSC is still poorly understood. Multifactorial genesis is presumed, including a proposed genetic susceptibility for PSC. 3 Genome-wide association studies have enabled researchers to look into a wide array of genetic variants to identify patterns of risk loci. Additionally, published data suggest that single nucleotide polymorphisms (SNPs) in PSC patients are associated with genes coding for certain phenotypes in cholangiocyte biology or tumour formation, and can also represent pivotal proteins in fibrogenesis and immunology, which includes IL-2-R or HLA-complex. 4 Stasis of bile is a known driver of inflammation and ultimately hepatic fibrosis. 5,6 During cholestasis, individual bile acids and other xenobiotic compounds present in bile expose biliary epithelial cells to a constant stimulus for inflammation, cellular stress and proliferation. 7,8 The UDP-glucuronosyltransferase(UGT)1A family of proteins plays a major role in the detoxification and elimination of bilirubin, bile acids and a broad array of potentially cytotoxic xenobiotics, which are linked to inflammation. 9 To date, the relationship between the functionally important UGT1A proteins and their naturally occurring genetic variants and the course and outcome of PSC has not yet been established.
The UGT1A gene locus encodes nine active protein isoforms, which are expressed in a tissue specific fashion in many organs. 10 Although UDP-glucuronosyltransferases are very abundantly expressed in the liver (eg UGT1A1 and UGT1A3), extrahepatic glucuronidation is specifically enabled by, for example, the exclusively extrahepatic UGT1A7 in the upper gastrointestinal tract. 11,12 Genetic variations of the UGT1A genes are frequent and alter the function of the affected genes. 13 Most prominently, polymorphisms of the UGT1A genes are known to cause hyperbilirubinaemia syndromes such as Gilbert-Meulengracht and Crigler-Najjar syndromes. 14 Other genetic variants are associated with the occurrence of different types of carcinomas, for example in bladder, breast, oesophagus, colorectal tract and lung. 10,15,16 The risk to develop cholangiocarcinoma (CCA) is significantly higher in patients with PSC. 17 Whether specific UGT1A variants are associated with a higher risk for CCA or contribute to this risk is presently unknown.
Clinical observation has shown that the phenotype and risk for colorectal malignancy in individuals with IBD in PSC have been found to differ from those without PSC. 18 A milder clinical course is usually observed in PSC-IBD 19 and bile acid and microbiota appear to differ from those of regular IBD patient, 20,21 indicating gut-liveraxis-related factors. Altered glucuronidation of bilirubin or xenobiotics would represent an intriguing hypothesis for the development of IBD in PSC.
Therefore, cholestasis and the associated toxicity of bile components represent an important factor for PSC progression to cirrhosis or CCA. UGT1A is a key process determining the elimination and detoxification of pro-inflammatory and carcinogenic compounds.
We thus analysed UGT1A variants in healthy blood donors and in patients with PSC to compare clinical presentation, outcome and association with IBD in carriers of UGT1A variants and wild-type genes in order to establish marker for the course and presentation of PSC.

Key points
Primary sclerosing cholangitis (PSC) is an enigmatic disease with an unknown cause, a lack of efficient medical therapy, and an often unpredictable course leading to liver failure and cancer in many cases. The present study identifies a role of anti-oxidative and protective metabolizing enzyme (UGT1A), which are linked to the course of the disease and indicate a role for presentation and outcome of PSC. These data may help to identify additional target for medical therapy aimed at increasing protection from PSC progression.

| Data collection
At study entry the following clinical characteristics at date of diagnosis of PSC were recorded from the patient charts: Biochemical blood tests

| Genotyping of UGT1A variants
Genomic DNA was isolated from whole blood samples by the

| Statistical analyses
Data were analysed using the SPSS 22.0 software package for Windows (SPSS Inc.). Continuous variables are presented as means or medians and were compared by Mann-Whitney U-test.
Categorical variables were compared by chi-squared test. The actuarial survival free of LT rate was estimated by Kaplan-Meier survival analysis. Patients were grouped according to their different UGT1A variants and the differences between the actuarial estimates were analysed using the log-rank test. All tests were two-tailed and a Pvalue of .05 or less was considered statistically significant. PSC-AIH-phenotype was diagnosed in 13.9% of the patients (see Table 1). For the outcome analysis of the PSC cohort, we excluded three patients with an unclear follow-up status so that 328 patients remained. The median follow-up time after first diagnosis of PSC was 13.7 ± 8.3 years (range 0-38.2 years). Owing to LT in 131 patients and death before LT in 21 patients, the median transplant-free survival time after PSC diagnosis was 14.9 ± 0.73 years.

| Biochemical parameters and UGT1A variants
Direct serum bilirubin of PSC cohort was compared between UGT wild types and variants. Patients carrying either a heterozygous or homozygous UGT1A1*28 variant showed significantly lower direct bilirubin levels than subjects with UGT1A1 wild type (P = .002; see Table 3). In contrast, patients showing at least one polymorphism in all of the three genes showed significantly higher bilirubin concentrations than wild type (P = .02). There was a trend towards higher bilirubin levels in patients carrying the UGT1A7 N129K/R131K variant compared to the UGT1A7 wild-type cohort (P = .089; see Table 3).
There Serum cholesterol was found to be higher in every tested polymorphism compared to wild type (see Table 3). 3.4 | UGT1A1, UGT1A3 and UGT1A7 wild types are associated with a higher incidence of IBD Ulcerative colitis and Crohn's disease were present significantly more often in patients carrying wild-type alleles for all three genes than in patients with at least one haplotype (P = .046) or homozygous variants (P = .022; see Figure 1). The least prevalent wild type of the three tested genes in patients without IBD was UGT1A7 (12.9%). Among patients with IBD, UGT1A7 wild type was more common in patients with ulcerative colitis (20.4%) than in cases of Crohn's disease (14.8%), but this difference was not significant (P = .436).

| Risk of CRC in PSC-IBD with homozygous UGT1A3-66T>C
In

| UGT1A variants are not associated with PSC-AIH-phenotype or IgG4 levels
About 13.9% of our patient cohort was diagnosed with PSC-AIHphenotype. Prevalence of UGT1A variants or wild types, respectively, did not differ between patients with concomitant AIH and patients suffering from PSC only (see Table S2). Transplant-free survival of patients with PSC-AIH-phenotype did not differ from the rest of the cohort (P = .21).
IgG4 levels at the beginning of follow-up were available in 153 cases. IgG4 levels did not differ between genotypes (see Table S2).
Patients who reached end point death or LT showed significantly higher IgG4 levels (P = .029). In multivariate Cox-regression analyses, IgG4 (P = .001) and UGT1A7 (P = .022) or UGT1A1 and 3 and 7 (P = .028) respectively proved to be independent risk factors for survival.
About 25.1% of the cases was found to suffer from other autoimmune diseases that are described above. Frequencies of autoimmune comorbidities were comparable between variants and wild types of UGT1A1 and 3 and 7 (see Table S2). Comorbidities were not associated with transplant-free survival (P = .44). 17.2 ± 3.7 years, P = .048; Figure 2

| Malignancy-free survival does not differ between genetic variants and wild type
Hepatobiliary malignancy, defined as cholangiocellular carcinoma (n = 23), gallbladder cancer (n = 5) or hepatocellular carcinoma (n = 1), developed in 8.76% of the PSC patients. The prevalence of these malignancies was not associated with any of the studied genotypes.

| D ISCUSS I ON
The UGT1A1*28 polymorphism is well established to cause one of the most frequent genetic syndromes, namely Gilbert disease by a reduction of bilirubin conjugation. This variant has been shown to be associated with a number of other additional UGT1A variants, which are also characterized by an altered enzymatic activity 24 and have been associated with cytotoxicity and genotoxicity as well as cancer disposition. However, in our initial assessment the overall prevalence of polymorphisms did not differ between healthy controls and our large cohort of patients with well-defined PSC, which is in line with findings of the recent genome-wide association studies, which identified various risk loci but did not point to the UGT1A gene locus on chromosome 11. However, in a more detailed analysis, UGT1A1*28 was associated with significantly lower levels of direct serum bilirubin at first diagnosis of the disease compared to wild type. The most likely explanation is a lower inflammatory activity leading to lower levels of conjugated bilirubin. UGT1A1*28 leads to fluctuating elevated levels of unconjugated (indirect) bilirubin (levels not given in the study), which has been described to lead to antioxidant activity, less inflammation and protection against diseases such as coronary heart disease. 25 The data indicate that UGT1A1*28 could be viewed as a protective genetic trait in PSC. But the more detailed analysis of a complex haplotype of variants also indicates that UGT1A variants are associated with an increased risk disposition.
Carriers of the UGT1A7 N129K/R131K allele showed a trend to higher conjugated bilirubin concentrations, likely reflecting a more severe clinical course of PSC. Therefore, because early LT and death represent clinical end points indicative of a more progressive course of PSC, we compared transplant-free survival. In this analysis carriers of at least one variant UGT1A allele are characterized by an inferior outcome that carriers of wild-type UGT1A alleles.
In combination, our data suggest that UGT1A variant is likely to exert a differential effect on the course of PSC, UGT1A7 N129K/R131K reflecting a negative marker regarding survival, and UGT1A1*28 associated with milder cholestasis.

F I G U R E 5
Kaplan-Meier estimates of cumulative transplant-free survival after PSC diagnosis for UGT1A3 wild-type carriers and patients with heterozygous or homozygous UGT1A3-66T>C alleles. Pvalue is calculated using log-rank test Apart from overall survival the association with IBD is a prominent feature of PSC, which may also be influenced by variations of glucuronidation. The overall analysis suggests that in the studied large cohort of PSC patients, for carriers of wild-type alleles of UGT1A1, UGT1A3 and UGT1A7, the risk for IBD is increased.
Interestingly, it has been reported previously that patients with Crohn's disease were less likely to be homozygous for UGT1A1*28 than healthy controls. 26  to the regulation of detoxification by glucuronidation. Currently, in many centres the medical therapy of PSC is mainly based on ursodeoxycholic acid (UDCA), a bile acid that is also glucuronidated by UGT1A3. 36 Other bile acids like nor-UDCA or obeticholic acid are under investigation as new treatment strategies. Against this background, our findings illustrate the need to further elucidate the role of UGT1A variants as disease modifier in PSC and the role of inducers not only of nuclear receptors but also of detoxification enzymes such as UGT1A proteins and their specific activators. 37

CO N FLI C T O F I NTE R E S T
The authors don't have any disclosure to report.